上海科技大学知识管理系统

Due to the vulnerability of the Internet of Things (IoT), it is indispensable to provide adequate security services. These services are generally implemented through security protocols or algorithms and running on energy-sensitive IoT devices. In order to design energy-efficient algorithms to prolong the lifetime of IoT devices, the energy characteristics of those algorithms should be analyzed primarily. In this paper, we conduct an integrated static analysis method with dynamic tracing to provide a quantitative energy profile for popular security algorithms such as AES128, RSA, and SHA256. Specifically, binary instructions executed in the invoked function are measured and counted through remotely debugging each program on the Arm development board. Then, the fine-grained energy consumption inside a program is revealed by combining the instruction statistics and instruction-level energy model. The experimental results show that the energy consumption of a program is mainly consumed by a few primary functions and CPU-memory interaction instructions, and hence the functions can be implemented in different ways to reduce energy consumption. This meaningful energy consumption evaluation method for security algorithms is able to guide to optimizing existing algorithms for embedded security.